In the past decade, many organizations implemented Enterprise Resource Planning (ERP) software to have an integrated, single software solution on which they could run their business. Client-server technology that was first introduced by the SAP R/3 ERP software became the de facto industry standard for business application software. Client-server technology separates the client (front end) from the application server that is running the application software. Further, the application server is separated from a database server that is running the database software. The database holds all data that are relevant to the application software. The high scalability of this multi-tier approach allowed huge numbers of users in an organization to work simultaneously in the same application system because the load of the system could be better balanced between application servers and database servers compared to systems where applications were run on the data base server.
Driven by the Internet and the idea to optimize business processes not only inside organizations, but also across organizations, many new business software applications were developed to complement the inner-organizational ERP application software. These additional software applications were designed to enable business processes across companies. Examples are Customer Relationship Management (CRM) applications to improve the management of customer relationships and Supply Chain Management (SCM) applications to improve logistics between companies and supplier relationships. However, this approach results in multiple, different and distributed software packages and increases the complexity of system landscapes in organizations.
The typical landscape of today consists of distributed systems with different software applications and separate servers. Each system requires its own database on its own physical hardware. The reasons for this are hardware and software related. For example, most symmetric multiprocessor (SMP) systems were limited to less than 16 processors and less than 4 GB of shared main memory. Operating systems did not offer the feature of creating multiple logical partitions for groups of processors. Databases only allowed for a maximum number of manageable tables (e.g. 10000).
FIG. 1 illustrates an example of an organization's computer system landscape of the prior art. The application systems ERP 1, CRM 2 and SCM 3 are connected through a network. The network is a local area network (LAN) or a wide area network or any other network known in the art and comprises the branches 10 and 11. Each of the application systems requires its own database system (DB) with the corresponding application data, for example, ERP DB 4, CRM DB 5 and SCM DB 6. Each database system 4-6 is connected to the corresponding application system 1-3 via network branches 12-14. Each application system 1-3 provides application services to pluralities 20-22 of application users. Multiple application users of each of the pluralities can be assigned to a single person. All application users of the pluralities can connect 23-25 to the corresponding application system 1-3 through, for example, a browser based front end computer (not shown) via the network. The system of the prior art, therefore, implies the need of providing three database server computers as well as the need of administrating (backup, tune, manage) these three different database systems.
Further, in case of a total or partial system break-down it is very difficult to have a consistent backup of all database systems. This is critical, because the different application software applications exchange data. For example: an order gets created by a sales person using an internet sales application in the CRM system 2. Then, the order is transferred into the order fulfillment application of the ERP system 1. The SCM system 3 analyzes the order for availability by using the available-to-promise application and finally an analytical application system for simulations (not shown) creates a forecast for the order. Therefore, a full and consistent point-in-time recovery of all databases in the system landscape is desired to guarantee data integrity of all business application data.